Choosing cathode bypass values

[Gnobuddy]

Awesome. Thanks.
What would you call something that goes viral to cure a 'virus'?

A meta-virus?

-Gnobuddy

[Gnobuddy]

I think it's more effective to switch the value of the coupling cap feeding the vol pot.

That looks like a nice scheme. As mentioned earlier in this thread, if (Rk) and (1/gm) are not too far apart in value, the cathode bypass cap really doesn't do a very good job of acting as a high-pass filter, because the two corners of the shelving filter are only a few dB apart.

-Gnobuddy

[Gnobuddy]

Here is the corrected script:

Your script doesn't seem to agree well with LTSpice at all - it wants a 10.6 uF cap to match what LTSpice shows with a 1 uF capacitor. See screenshots.

In cases where (1/gm) is close to (Rk), there may never be a (-3dB) frequency at all - response goes all the way down to zero hertz without dropping 3 dB from the fully-bypassed frequencies.

I believe using (rp) is of marginal importance in most cases - I don't know of a valve where (rp) is low enough to be comparable to (1/gm). Perhaps this might occur in some ancient, low-transconductance, high-power triode output valve?

-Gnobuddy

[nickb]

Your script doesn't seem to agree well with LTSpice at all - it wants a 10.6 uF cap to match what LTSpice shows with a 1 uF capacitor. See screenshots.

In cases where (1/gm) is close to (Rk), there may never be a (-3dB) frequency at all - response goes all the way down to zero hertz without dropping 3 dB from the fully-bypassed frequencies.

I believe using (rp) is of marginal importance in most cases - I don't know of a valve where (rp) is low enough to be comparable to (1/gm). Perhaps this might occur in some ancient, low-transconductance, high-power triode output valve?

-Gnobuddy

You entered the wrong data.



PS: The case where you can't get -3dB is detected and reported.

[Gnobuddy]

You entered the wrong data.

Ah. You removed my "gm" entry, and replaced it with "rp" and "mu", which can be used to calculate gm.

Nice calc! For your next trick, add display of the frequency response - easy enough to do with an HTML canvas element, but other things are calling me more urgently.

-Gnobuddy

[g1]

Ah. You removed my "gm" entry, and replaced it with "rp" and "mu", which can be used to calculate gm.

I thought it was your 1M5 cathode resistor.

[Gnobuddy]

I thought it was your 1M5 cathode resistor.

Apparently my brain took a short vacation without telling me when it was leaving.

-Gnobuddy

[Chuck H]

Welcome Gnobuddy! Stick around. I think you'll fit right in

[nosaj]

Welcome Gnobuddy! Stick around. I think you'll fit right in

Some reason when I read your post Chuck David Lee Roth Chimed in singing I ain't got Gnobuddy.

nosaj

Some days our brains are on it an other days they have left the building.

[Malcolm Irving]

Going back to the various formulae. (Sorry to be boring! )

... However, the formula itself, while very popular, is actually wrong. The "R" that matters isn't the external cathode resistor. It's the invisible "internal" cathode resistor, which has a value of (1/gm).

More accurately, the proper resistance to use is the internal (1/gm) resistance in parallel with the external cathode resistance Rk. Usually (but not always!), the parallel combination is dominated by the (1/gm) term, as Rk is usually much larger than (1/gm). So usually we can ignore Rk, and just use (1/gm). ...

Merlin Blencowe (in Designing High-Fidelity Tube Preamps, 2016) gives the pole frequency as:

1 / ( 2.pi.Ck ( Rk || rk ) )

Where rk = (Ra + ra)/mu

It is well known that mu = gm.ra and consequently 1/gm = ra/mu

rk is not equal to 1/gm

but rather rk = Ra/mu + 1/gm

[nickb]

Ah. You removed my "gm" entry, and replaced it with "rp" and "mu", which can be used to calculate gm.

Nice calc! For your next trick, add display of the frequency response - easy enough to do with an HTML canvas element, but other things are calling me more urgently.

-Gnobuddy

Yup. Wanted to stick to the RDH4 formula, to avoid confusion, funnily enough. Sorry about that.

Charts? Done! Give it a go Cap Calculator

[Thoriated Tungsten]

Yup. Wanted to stick to the RDH4 formula, to avoid confusion, funnily enough. Sorry about that.

Charts? Done! Give it a go Cap Calculator

There seem to be some small buglets.

You cannot enter values, which are not a multiple of 10. For instance mu of 41 or ra of below 10k. (Tried to use ra of 8.3k, as an example). Mu below 10 is also a no-go for the same reason. Also, there are a few small typos.

(2) Total effective plate load re_istance, k ohms - missing 's' in the indicated position.
(4) Cathode resistor, k ohms) - One ')' too many.

[nickb]

There seem to be some small buglets.

You cannot enter values, which are not a multiple of 10. For instance mu of 41 or ra of below 10k. (Tried to use ra of 8.3k, as an example). Mu below 10 is also a no-go for the same reason. Also, there are a few small typos.

(2) Total effective plate load re_istance, k ohms - missing 's' in the indicated position.
(4) Cathode resistor, k ohms) - One ')' too many.

Thx for the feedback . The things you saw as buglets were design choices. I've changed them where I felt appropriate.

[Malcolm Irving]

Using the default values in Nickb’s on-line calculator (post #41): mu=100, Rp(total)=95k, rp=63k, Rk=1.5k and a desired -3dB frequency of 150Hz, we get Ck=0.96uF

Substituting these values into Merlin’s formula for the pole frequency (see post #40) we get a pole frequency of 215Hz.

At first sight there appears to be a discrepancy. However, I think it might be that the pole frequency is not an exact predictor of the -3dB point, when there is a zero close by.

[nickb]

Using the default values in Nickb’s on-line calculator (post #41): mu=100, Rp(total)=95k, rp=63k, Rk=1.5k and a desired -3dB frequency of 150Hz, we get Ck=0.96uF

Substituting these values into Merlin’s formula for the pole frequency (see post #40) we get a pole frequency of 215Hz.

At first sight there appears to be a discrepancy. However, I think it might be that the pole frequency is not an exact predictor of the -3dB point, when there is a zero close by.

I believe that is precisely the reason for the difference. The pole and zero are very close and so interact with one another.